meitu-beauty

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt explicitly instructs configuring AK/SK via commands like meitu config set-ak --value "..." (CLI args containing secrets) which would require the agent to include user-provided secret values verbatim, posing an exfiltration risk.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.80). The skill explicitly accepts image URLs in the "Execute → 输入获取" step and will fetch and inspect/process those external images (face detection, single-person checks, error-handling and retries), so untrusted third-party image content from the open web can directly influence processing decisions and tool actions.